Directed graphs may be used to model information technology (IT) infrastructures. As an IT infrastructure expands, it may be more difficult for a graph modeling the infrastructure to be searched. Furthermore, applications that integrate with the graph may not be interested some information contained in the graph. For example, a management application may list computers at various locations, but may not use data describing routers and switches to which the computers are connected, even though the IT infrastructure graph may contain networking information. In another example, applications that integrate with the graph may be interested in information which is not directly stored in the graph. Conventionally, when an application desires additional information, the graph may be modified to contain this information. However, this may not be efficient.
Typical applications are designed to be integrated with specific graph models and to handle unnecessary information at the application level. This may include removing or ignoring unnecessary information when it is acquired from the graph model, as well as adding information otherwise unimportant to the application when sending information to the graph model. However, as the number of applications increases over time, it may not be economical for applications to be designed to work with specific models. Furthermore, when the graph model is updated to a new version, applications that are not configured to work with the new version may cease to function.